First Step Towards End-to-End Parametric TTS Synthesis: Generating Spectral Parameters with Neural Attention

By Wenfu Wang, Shuang Xu, Bo Xu. 2016

61 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Tao Qin
2. Xu Tan
3. Tie-Yan Liu
4. Frank K. Soong
5. Lei He
6. Lei Xie
7. Yu Zhang
8. Dong Yu

In Signal Processing

1. Frank K. Soong
2. Xu Tan
3. Tao Qin
4. Tie-Yan Liu
5. Lei He
6. Dong Yu
7. Jinyu Li
8. Lei Xie

Citations in the corpus

• UniAudio: An Audio Foundation Model Toward Universal Audio Generation (2023, methodology): “…and phonetics (Zen et al., 2013), signal processing (Griffin & Lim, 1984), auditory perception (Shadle & Damper, 2001) and machine learning (Wang et al., 2016) research etc; (4) they use different generative model, such as diffusion model (Huang et al., 2023a; Shen et al., 2023; Wang et…”
• A Survey on Neural Speech Synthesis (2021, methodology): “From 2010s, neural network-based speech synthesis [419, 278, 76, 417, 369, 187, 248, 376] has gradually become the dominant methods and achieved much better voice quality.”
• Tacotron-Based Acoustic Model Using Phoneme Alignment for Practical Neural Text-to-Speech Systems (2019, methodology): “• Many seq2seq methods introduce phoneme input [14–16,19, 23–35] rather than character input [10–22].”
• Tacotron: A Fully End-to-End Text-To-Speech Synthesis Model (2017, methodology): “To our knowledge, Wang et al. (2016) is the earliest work touching end-to-end TTS using seq2seq with attention.”
• A Review of Deep Learning Based Speech Synthesis (2019, methodology): “For example, the authors in [55] employed the structure with content-based attention [56] to model the acoustic features for speech synthesis.”
• Forward Attention in Sequence- To-Sequence Acoustic Modeling for Speech Synthesis (2018, unknown): “To our knowledge, the first work among them [18] adopted content-based attention [12] to build the encoder-decoder acoustic model for speech synthesis.”
• On the localness modeling for the self-attention based end-to-end speech synthesis (2020, methodology): “…approach typically conducts frame-level regression modeling, so we need an extra module to align texts and corresponding acoustic features in time Xu, and Xu (2016), the authors firstly proposed an end-to-end TTS framework with an attention-based seq2seq model, but an extra HMM-based aligner is…”
• Improving Prosody with Linguistic and Bert Derived Features in Multi-Speaker Based Mandarin Chinese Neural TTS (2020, methodology): “This framework first achieved great success in machine translation [7], and then was applied to speech recognition [8][9] and synthesis [10][6][11].”
• Implementing Prosodic Phrasing in Chinese End-to-end Speech Synthesis (2019, methodology): “When the seq2seq model with attention was first introduced for speech synthesis, a rough alignment was needed for the method to work [9].”